ISYE 6414 - Summer 2009 Solution Lecture 9. Problem 2 1. If c = 0, ˆ β R is equal to ˆ β * , that is, the estimators for the standarized data. 2. If c → ∞ all ˆ β R k goes to zero. Problem 3 Given that Cov ( x k ,x l ) = 0, the components can be chosen as T 1 = x * 1 , where x * 1 is the predictor with the highest correlation with the response Y . Then, at each iteration (i-th), the matrix of centered predictors is reduced by the variables already used, and
This is the end of the preview. Sign up
access the rest of the document.